Fordham University Masthead Logo DigitalResearch@Fordham

Chemistry Faculty Publications Chemistry

1986 The ubs cellular localization of glutamate dehydrogenase (gdh): is gdh a marker for mitochondria in brain? / James C. K. Lai, Kwan-Fu Rex Sheu, Young Tai Kim, Donald D. Clarke, and John P. Blass Department of Neurology, Cornell University Medical College and Altschul Laboratory for Dementia Research Burke Rehabilitation Center White Plains, NY 10605 and Department of Medicine Cornell University Medical College New York, NY 10021 James C. K. Lai Cornell University. Department of Neurology and Neuroscience

Kwan-Fu Rex Sheu Burke Rehabilitation Center

Recommended Citation Lai, James C. K.; Sheu, Kwan-Fu Rex; Kim, Young Tai; and Clarke, Donald Dudley PhD, "The ubcs ellular localization of glutamate dehydrogenase (gdh): is gdh a marker for mitochondria in brain? / James C. K. Lai, Kwan-Fu Rex Sheu, Young Tai Kim, Donald D. Clarke, and John P. Blass Department of Neurology, Cornell University Medical College and Altschul Laboratory for Dementia Research Burke Rehabilitation Center White Plains, NY 10605 and Department of Medicine Cornell University Medical College New York, NY 10021" (1986). Chemistry Faculty Publications. 21. https://fordham.bepress.com/chem_facultypubs/21

This Article is brought to you for free and open access by the Chemistry at DigitalResearch@Fordham. It has been accepted for inclusion in Chemistry Faculty Publications by an authorized administrator of DigitalResearch@Fordham. For more information, please contact [email protected]. Young Tai Kim Cornell University. Medical College

Donald Dudley Clarke PhD Fordham University, [email protected]

Follow this and additional works at: https://fordham.bepress.com/chem_facultypubs Part of the Biochemistry Commons • Neurochemical Research, Vol. 11, No.5, 1986, pp. 733-744

THE SUBCELLULAR LOCALIZATION OF GLUTAMATE DEHYDROGENASE (GDH): Is GDH a Marker for Mitochondria in Brain?

JAMES C. K. LAI, KwAN-Fu REx SHED, YouNG TAI KIM*, DoNALD D. CLARKE, and, JoHN P. BLAss Department of Neurology, Cornell University Medical College and Altschul Laboratory for Dementia Research Burke Rehabilitation Center White Plains, NY 10605 and *Department of Medicine Cornell University Medical College New York, NY 10021

Accepted December 3, 1985

Glutamate dehydrogenase (GDH, EC 1.4.1.2) has long been used as a marker for mitochondria in brain and other tissues, despite reports indicating that GDH is also present in nuclei of liver and dorsal root ganglia. To examine whether GDH can be used as a marker to differentiate between mitochondria and nuclei in the brain, we have measured GDH by enzymatic activity and on immunoblots in rat brain mitochondria and nuclei which were highly enriched by density-gradient centrifugation methods. The activity of GDH was enriched in the nuclear fraction as well as in the mitochondrial fraction, while the activities of other "mitochon­ drial" enzymes (fumarase, NAD-isocitrate dehydrogenase and pyruvate dehy­ drogenase complex) were enriched only in the mitochondrial fraction. lmmunob­ lots using polyclonal antibodies against bovine liver GDH confmned the presence of GDH in the rat brain nuclear and mitochondrial fractions. The GDH in these two subcellular fractions had a very similar molecular weight of 56,000 daltons. The mitochondrial and nuclear GDH differed, however, in their susceptibility to solubilization by detergents and salts. The mitochondrial GDH could be solubi-

Please address reprint requests to: Dr. Sheu, Burke Rehabilitation Center, 785 Mamaroneck Avenue, White Plains, NY 10605. Dr. Lai's current address is: Department of Neurology, Cornell University Medical College, 1300 York Avenue, New York, NY 10021. Dr. Clarke's current address is: Chemistry De­ partment, Fordham University, Bronx, NY 10458. 733 0364-3190/86/0500-0733$05.00/0 © 1986 Plenum Publishing Corporation 734 LAI ET AL.

lized by extraction with low concentrations of detergents (0.1% Triton X-100 and 0.1% Lubrol PX), while the nuclear GDH could be solubilized only by elevated concentrations of detergents (0.3% each) plus KCl (> 150 mM). Our results indicate that GDH is present in both nuclei and mitochondria in rat brain. The notion that GDH may serve as a marker for mitochondria needs to be re-evaluated.

INTRODUCTION

Glutamate dehydrogenase [GDH, EC 1.4.1.2, L-glutamate: NAD oxido­ reductase (deaminating)] has long been thought to be confmed to mito­ chondria (1), and has been used extensively as a marker to identify mi­ tochondria in subcellular fractions of many tissues, including the brain. However, the inference that the GDH activity can be used as a measure of mitochondrial contamination in subcellular fractions other than mito­ chondria may not be justified. Di Prisco and his coworkers have purified GDH from isolated nuclei of bovine liver (2). This nuclear GDH was not due to a contamination of mitochondrial origin since the immunochem­ ical, catalytic and chromatographic properties of this nuclear GDH were not identical, although very similar, to those of mitochondrial GDH (2, 3). Kato and Lowry (4) have also found GDH activity in the nuclei dis­ sected from single dorsal root ganglion neurons of the rabbit. The GDH activity persisted in the trimmed and subdivided nucleus, and accounted for i the specific activity (on a dry weight basis) relative to the rest of the cell (4). The brain consists of many different types of cells. In accord with this, variation of GDH has been reported at the regional level (5, 6), and is likely to be manifested at the cellular and subcellular levels (7 -9). A significant amount of GDH confined to the nuclei may result in a mis­ judgment of the purity of isolated nuclear fractions if GDH is employed as a marker for mitochondria. In this study, we determined whether there is a significant amount of GDH in the isolated brain nuclei by using activity measurement and im­ munoblotting techniques. The results support the long overlooked notion that GDH is also present in nuclei, and that GDH may not be a proper marker for mitochondria, especially in brain tissue. Preliminary results of this study have been presented (10, 11).

EXPERIMENTAL PROCEDURE

Preparation of Nuclei. Four male Wistar rats (Charles River Breeding Laboratories, Sto­ neridge, NY) of 45-60 days old were used in each experiment. The forebrains were dissected out as described previously (12), the superficial blood vessels removed, and the brains rinsed once with the ice-cold isolation medium which contained: 0.32 M sucrose, 1 mM MgCh, • SUBCELLULAR LOCALIZATION OF GDH 735

and 1 mM 3-(N-morpholino)propanesulfonic acid, pH adjusted to 6.6 with Tris base [MOPS(Tris), pH 6.6]. The brains were then minced with scisssors, and further rinsed at least 3 times with the isolation medium. The chopped tissue was manually homogenized with 9-volumes of isolation medium in an all glass Dounce homogenizer with loose pestle (clearance: 0.007 inch) using 10 up- and down-strokes. The homogenate was filtered twice through 110 m stainless-steel meshes to remove cellular debris and capillaries. The filtrate was then centrifuged at 3,000 g for 3 min. The supernatant and the loose portion of the pellet were decanted into another centrifuge tube. The tightly packed pellet, which consisted mostly of capillaries, some cellular debris and portions of large nuclei, was discarded. The decanted material was then centrifuged at 3,000 g for 3 min. The pellet, which consisted mainly of nuclei, was resuspended in 30 ml of 2 M sucrose, 1 mM MgCh and 1 mM MOPS(Tris), pH 6.6, and divided into two 30 ml centrifuge tubes. To each tube, 15 ml of 1 M sucrose, 1 mM MgCh and 1 mM MOPS(Tris), pH 6.6 was carefully layered on top of the suspension which contained nuclei. The gradient was centrifuged at 110,000 g for 1 hr in a SW-27 rotor in a Beckman L5-50B ultracentrifuge. The nuclear pellet was then washed once with the isolation medium, suspended in the isolation medium (10 mgp/ml) and stored at - sooc in small aliquots. The yield was 5.9 + 0.9 mgp/g wet weight (mean + SD; n = 7). Preparation of Mitochondria and Cytosol. Male Wistar rats of similar ages (see above) were used in these preparations. Non-synaptic mitochondria from forebrains were isolated by the Ficoll-sucrose density gradient method, as detailed previously (13). The postmito­ chondrial supernatant was further centrifuged at 80,000 g for 90 min to obtain the cytosolic fraction. The homogenate, cytosol and mitochondria [resuspended 10 mgp/ml in 0.32 M sucrose and 5 mM N-2-hydroxyethylpiperazine-N' -2' -ethanesulfonic acid, pH adjusted to 7.4 with KOH] were also stored at - 80°C in small aliquots. Enzyme ~ctivity Measurement. The GDH activity was determined by the 2-oxoglutarate-, NHt -, and ADP-dependent NADH oxidation, as described by Lai and Clark (14). The pyruvate dehydrogenase complex (PDHC, EC 1.2.4.1, EC 2.3.1.12, and EC 1.6.4.3) was measured according to Sheu et al. (15). The NAD-linked isocitrate dehydrogenase (ICDH, EC 1.1.1.41) and fumarase (FUM, EC 4.2.1.2) were measured according to Lai and Clark (14). These enzyme activities were measured within a time period when they remained fully active: the ICDH and PDHC were assayed within 2 days, and the GDH and fumarase within a week. Antibodies Against GDH (Anti-GDH). The bovine liver GDH obtained from Sigma Chem­ ical Co. (St. Louis, MO) was further purified by sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The GDH was electrophorased in 7.5% polyacrylamide slab gels (Hoefer 500SE unit) according to Weber and Osborn (16), and briefly stained with Coomassie blue and destained. The GDH band was cut out from the gel slabs, and embedded with 6% polyacrylamide gel in the second 1.5 em gel tubes. The bottom ends of these gel tubes were sealed with dialysis bags. The GDH was then electrophoresed (Hoefer GT unit) into the dialysis tubes and concentrated there. This purified GDH (500 tJ.g protein) was emulsified with Freund's complete adjuvant, and injected subcutaneously into a 2-month old New Zealand white rabbit. Boosters of 500 f..Lg GDH were administered every 4 weeks afterwards. The rabbit was bled one week after each booster, and the anti-GDH titer in the serum was tested with the enzyme-linked immunoabsorbent assay (see below). Antibodies specific for GDH were detected at a 2,000-fold dilution of the serum after the first booster. The -y-immunoglobulin (IgG) fraction from the pooled sera was prepared by (NH4hS04 precipitation at 50% saturation twice, and DEAE-cellulose (Bio-rad Laboratories, Rich­ mond, CA) column chromatography (17). Immunochemical Assays for GDH. An enzyme-linked immunoabsorbent assay was em­ ployed to detect anti-GDH in the immune serum as described earlier (17). Aliquots of 200 736 LAI ET AL. •

J,Ll of bovine liver GDH antigen (5 J,Lgp/ml, in 100 mM Na-carbonate, pH 9.6) were coated to the wells of a 96-well polyvinyl microplate. The wells were subsequently blocked with bovine serum albumin, incubated with serial dilutions of the immune serum or control serum. The immunoglobulins that bound to the GDH antigen were detected with the alkaline phos­ phatase conjugated goat-anti-rabbit lgG second antibodies (Dynatech Laboratories, Alex­ andria, VA) and p-nitrophenol phosphate. The yellow color of p-nitrophenol was read on a Bio-tek EL307 EIA reader. The GDH in the brain preparations was also analyzed by immunoblotting techniques. The brain homogenate or subcellular fractions were resolved by SDS-PAGE (7.5% gel). The resolved polypeptides were then transfered electrophoreticaly to a nitrocellulose membrane (Bio-rad) in a modified Pharmacia GD-4 gel destainer, as described previously (18). To visualize the GDH-immunocross-reacting material, the nitrocellulose transfer was blocked with the bovine serum albumin, and incubated with the 200-fold diluted anti-GDH lgG (0.02 mgp/ml), horseradish peroxidase conjugated goat-anti-rabbit lgG second antibodies (Bio­ rad), and 4-chloro-1-napthol and H20 2 , as detailed previously (18). The GDH-immunocross­ reacting material appeared as the peroxidase-stained blue bands. Other Procedures. The protein concentration was determined according to Lowry et al. (19) using bovine serum albumin as the standard. Statistical significance was evaluated by Student's two-tailed t test.

RESULTS

Nuclei were isolated from rat brain in buffered sucrose solutions (pH 6~6) containing 1 mM MgCh (20) by differential centrifugation, and dis­ continuous sucrose density-gradient centrifugation (Experimental Pro­ cedure) which was designed to obtain a population of nuclei without the contamination of mitochondria. The brain nuclei isolated by similar su­ crose-density gradient procedure were pure on electron microscopic ex­ amination (20-22; Lai, J. C. K., unpublished observation). Mitochondria of non-synaptic origin were also isolated by a Ficoll-sucrose density gra­ dient method (13). The activity of GDH; as well as the activities of fu­ marase (FUM), NAD-linked isocitrate dehydrogenase (ICDH) and py­ ruvate dehydrogenase complex (PDHC) which have been regarded as "mitochondrial" maker enzymes, were measured in these isolated mi­ tochondrial and nuclear fractions, and in the homogenate and the cytosolic fraction (Table I). As expected from their major mitochondrial localiza­ tion, the specific activities of all these enzymes were higher in the mi­ tochondrial fractions, but very much lower in the cytosolic fraction, when compared to those in the homogenate. On the other hand, activities of all these 4 enzymes were detectable in the nuclear fraction. Although specific activities of FUM, ICDH, and PDHC in the nuclear fraction did not appear to differ significantly from those in the homogenate, the spe­ cific activity of GDH was significantly higher in the nuclear fraction than in the homogenate by as much as 50%. Moreover, the activity ratios of • SUBCELLULAR LOCALIZATION OF GDH 737

TABLE I i ENZYME AcTIVITIES AND RATIOS OF GDH TO OTHER MITOCHONDRIAL ENZYME AcTIVITIES IN RAT BRAIN HoMOGENATE AND SuBCELLULAR FRACTIONs

Homogenate Mitochondria Nuclei Cytosol

Enzyme Activities GDH 205 + 42 (7) 699 + 150 (3) 307 + 26 (7)+ 3.9 + 0.2 (3) FUM 234 + 39 (7) 781 + 115 (3) 210 + 42 (7) 25.4 + 0.6 (3) ICDH 44.3 + 12.4 (7) 153 + 24 (3) 34.4 + 8.8 (7) 0.3 + 0.1 (3) PDHC 24.7 + 2.7 (6) 116 + 20 (3) 33.1 + 12.3 (5) N.D. Ratios of Enzyme Activities GDH/FUM 0.87 + 0.07 (7) 0.89 + 0.13 (3) 1.42 + 0.11 (7)** GDH/ICDH 5.24 + 0.59 (7) 4.51 + 0.50 (3) 7.74 + 1.80 (7)* GDH/PDHC 7.38 + 0.56 (5) 6.00 + 0.81 (3) 10.1 + 2.0 (5)*

Values are mean + SD with the number of experiments in parentheses. Enzyme activities are nmol/min/mg protein. N.D., not detectable. Abbreviations are GDH, glutamate dehy­ drogenase; FUM, fumarase; ICDH, NAD-linked isocitrate dehydrogenase; and PDHC, py­ ruvate dehydrogenase complex. Significantly different vs homogenate and mitochondria: *, P < 0.01; **, P < 0.001. Significantly higher than homogenate: +, P < 0.001.

GDH to FUM, GDH to ICDH and GDH to PDHC in the nuclear fraction were substantially higher than the corresponding ratios in the homogenate or in the mitochondrial fraction. These results indicate that the GDH is specifically enriched in the nuclear fraction among these 4 enzymes con­ ventionally regarded as "mitochondrial" markers. Based on other evi­ dence (see Discussion) that GDH was also found in nuclei from other tissues, the most likely interpretation for this result is that GDH is also present in brain nuclei. The effect of detergents (Triton X-100 plus Lubrol PX) and salt (KCI) on the solubility of GDH in the mitochondrial and nuclear fractions were compared as shown in Table II. The GDH in the mitochondrial fraction could be more eeasily solubilized than the GDH in the nuclear fraction. Treatment with 0.1% Triton X-100 plus 0.1% Lubrol PX, followed by freezing and thawing completely solubilized the GDH in the mitochondrial fraction whereas less than half of the GDH in the nuclear fraction was sol­ ubilized. The nuclear GDH was completely solubilized only in elevated concentrations of detergents (0.3% Triton X-100 and 0.3% Lubrol PX) and in the presence of KCl at concentrations higher than 150 mM (Table II). This result further indicates that the GDH associated with the nuclear fraction cannot be attributed entirely to a contamination of mitochondria, mitochondrial fragments, or other membraneous structures that enclose 738 LAI ET AL. •

TABLE II EFFECT OF DETERGENTS AND KCl ON THE SOLUBILITY OF MITOCHONDRIAL AND NucLEAR GLUTAMATE DEHYDROGENASE

% GDH activity solubilized Triton X-100 Lubrol-PX KCl (% (w/v)) (mM) Nuclei Mitochondria

0.1 0.1 0 42 104 0.3 0.3 0 37 92 0.3 0.3 50 63 105 0.3 0.3 75 83 109 0.3 0.3 100 88 109 0.3 0.3 150 97 109 0.3 0.3 300 109 0 0 80 13 53

Aliquots of mitochondria or nuclei were suspended in 5-volumes of 0.5 mM dithiothreitol, and detergents and KCl with final concentrations as indicated, frozen in liquid N2 for 2 min, and then thawed at 30°C. The GDH activities in these suspensions and in the supernatants following centrifugation at 9,600 g for 10 min were determined. Values are mean of 2 experiments.

mitochondria (e.g., synaptosomes, intact cells). It is also unlikely that the GDH associated with the nuclear fraction could be due to non-specific adsorption of the GDH that leaked out from mitochondria during the homogenization procedure since washing the nuclei with 80 mM KCl, or 0.1% Triton X-100 plus 0.1% Lubrol PX did not completely solubilize the GDH in the nuclear fraction (Table II). The mitochondrial and nuclear GDH were further compared immu­ nochemically. Rabbit antibodies were raised against the highly purified bovine liver GDH. Specific antibodies were obtained, as judged by the enzyme-linked immunoabsorbent assay (Figure-1A). The GDH antigen and the brain homogenate were also resolved on SDS-polyacrylamide gels by electrophoresis (16). On the nitrocellulose blots (Figure lB), 10 ng of bovine liver GDH antigen could be readily detected using the anti-GDH lgG. These antibodies recognized the rat brain GDH since the rat brain homogenate gave rise to one immunocross-reacting species over the range of 5 through 50 f..Lg protein tested (lanes 4-6, Figure lB). Substituting the anti-GDH with normal rabbit lgG gave rise to no visible bands (not shown). Furthermore, the GDH-immunocross-reacting species in rat brain had a similar molecular weight (56,000 daltons) when compared to that of the bovine liver GDH antigen (Figure lB). These observations • SUBCELLULAR LOCALIZATION OF GDH 739 A B

1 2 3 4 5 6

~ E c ..o.3 -• ,-• ,Z.2 ..-c ..c 0 at ,-c .1 c II-

.3 1 3 10 3 .. Dilution of Antlaerum (XI 000)

FIG. 1A (left panel). The enzyme-linked immunoabsorbent assay for antiserum for GDH. Serial dilutions of antiserum for GDH (e) and control serum (0) were added to microtiter plate wells which were pre-coated with bovine liver GDH antigen. The binding of anti-GDH was measured using the alkaline phosphatase conjugated goat-anti-rabbit IgG second anti­ bodies and p-nitrophenol phosphate, as outlined in Experimental Procedure. FIG. 1B (right panel). The immunoblot for GDH. Bovine liver GDH antigen (lanes 1-3, with 10, 30 and 100 ng, respectively) and rat brain homogenate (lanes 4-6, with 5, 15 and 50 f..Lgp , respectively) were resolved by SDS-PAGE. The GDH-immunocross-reacting material was identified on the nitrocellulose transfer using the anti-GDH IgG, horseradish peroxidase conjugated goat-antirabbit IgG second antibodies, and 4-chloro-1-napthol and H 20 2 , as out­ lined in Experimental Procedure.

indicated that this GDH-immunocross-reacting band reflected the au­ thentic GDH in rat brain. The brain homogenate and the subcellular fractions were analyzed for GDH on immunoblots (Figure 2). The GDH peptide could be detected in the homogenate, and the mitochondrial and nuclear fractions where GDH activity was present (Table 1). There was no visible band corresponding to GDH in the cytosolic fraction in which the GDH activity was very low (Table 1). Furthermore, the GDH in the mitochondria and nuclear frac­ tions had similar sizes, since a mixture of nuclear and mitochondrial frac­ tions showed no separation of GDH on the immunoblots {lane M + N, Figure 2). 740 LAI ET AL.

H C M N M•N

FIG. 2. Immunoblot for GDH in mitochondria and nuclei from rat brain. Brain homogenate and subcellular fractions were separated in the SDS­ polyacrylamide gel (9%) by electrophoresis. The GDH was visualized on the nitrocellulose transfer by immunoblotting technique, as out­ lined in Experimental Procedure. H, homoge­ nate, 6 ~gp; C, cytosolic fraction, 6 ~gp; M, mitochondrial fraction, 2 ~gp; N, nuclear frac­ tion, 6 ~gp; and M + N, mixture of the mito­ chondrial (2 ~gp) and nuclear (6 ~gp) fractions.

DISCUSSION

In this study, evidence is presented that GDH is present in nuclei as well as in mitochondria in brain. In the subcellular fractions of rat brain, GDH is enriched in the mitochondrial and nuclear fractions when com­ pared to the homogenate, while other "mitochondrial" enzymes (i.e., FUM, ICDH and PDHC) are enriched only in the mitochondrial fraction (Table I). Di Prisco and Casola (3) have shown that antibodies raised against the mitochondrial and nuclear GDH from bovine liver cross-re­ acted with the GDH from nuclei and mitochondria, respectively. The presence of GDH in the rat brain nuclear fraction was, therefore, iden­ tified independently on immunoblots using antibodies against the bovine liver GDH. The immunoblots also showed that the mitochondrial and nuclear GDH have a very similar molecular weight (56,000 daltons), a size reported by Chee et al. (23) of the highly purified rat brain GDH. Furthermore, our result is also consistent with the finding of King and Frieden (24) that the GDH in rat liver is apparently uniform with respect to molecular weight. Evidence in the literature has indicated that GDH is present in mito­ chondria as well as in nuclei from bovine (2, 3), rat (25), and pig (26) liver, • SUBCELLULAR LOCALIZATION OF GDH 741

in cultured hepatoma from human (27), and in dorsal root ganglion cells from rabbit (4). Except in the last study (4), all these investigations in­ volved the isolation of nuclei from liver cells using bulk-preparation meth­ ods. The conclusions of such studies have been challenged on the ground that the GDH in these nuclear fractions represents an artifact due to a contamination of mitochondrial origin (24). In the present study, a dis­ continuous sucrose density-gradient centrifugation method was used to isolate a population of nuclei free from mitochondria, synaptosomes and other membraneous fragments. Similar procedures have been used suc­ cessfully to isolate "pure" brain nuclei as determined by electron mi­ croscopic examination (20-22). A previous study based on electron mi­ croscopic examinations (Lai, J. C. K., unpublished observations) revealed that the mitochondrial contamination in the nuclear fraction iso­ lated by this procedure was significantly less than 5%, as determined by counting the number of mitochondria and nuclei. Since a nucleus is sig­ nificantly larger than a mitochondrion (c.f. 20-22), the extent of mito­ chondrial contamination in these nuclear preparations, on a per mg protein basis, was likely to be even less. The difference in the susceptibility of GDH in the mitochondrial and nuclear fractions to extraction by deter­ gents and salts (Table II) also supports the notion that the nuclear GDH cannot be derived entirely from mitochondria, or from intact cells, or other membraneous structure that enclose mitochondria, or from non­ specific adsorption of the GDH leaked out from mitochondria during the isolation process. Moreover, comparatively little GDH activity is detected in the cytosol (Table I and Figure 2), suggesting that the leakage of GDH from mitochondria is insignificant in comparison with the GDH found in the nuclei (Table 1). The low activities of FUM, ICDH and PDHC in the cytosolic fraction further support the notion that the isolated mitochondria are relatively intact. The resistance to solubilization of the nuclear GDH by detergents and KCl (Table II) is also consistent with the result of Kato and Lowry (4) who showed that the nuclear GDH is localized within the confines of the nuclear envelope. Immunohistochemical studies are cur­ rently underway to obtain additional information regarding the subcellular localization of brain GDH. In conclusion, the presence of GDH in nuclei indicates that the con­ ventional view that GDH is a specific marker for mitochondria needs to be re-evaluated. Other "mitochondrial" enzymes measured in this study (i.e., FUM, ICDH and PDHC) appear to be more specific in differentiating nuclei from mitochondria than does GDH in brain (Table 1). Due to a lack of information as to whether other mitochondrial enzymes are also present in nuclei (c.f. 4), it is advisable to use multiple enzyme markers, or in 742 LAI ET AL. • combination with other criteria, to identify the subcellular fractions of brain or to assess the purity of these fractions. Recently, a deficit of GDH has been reported in some patients with olivopontocerebellar atrophy (28-30) and with progressive supranuclear palsy (31). Plaitakis et al. (30) have shown that in leukocytes the deficit ofGDH is due to a deficit of the ''particulate'' GDH that remains insoluble after the leukocytes are treated with Triton X -100. Based on the difference in susceptibility to detergent extraction and to heat inactivation, it was proposed that these 2 forms of GDH represent isozymes of GDH (30). The data presented in this paper argue that the difference may (also) relate to the subcellular localization of the deficient GDH. Further characteri­ zation of the structure of these various forms of GDH is needed to clarify whether there exist isozymes of GDH, and to elucidate the pathophysi­ ological nature of the GDH-deficient neurological disorders. An immu­ noblotting procedure, similar to the one described in this study, has al­ ready been applied to identify the GDH in human tissues (31). Immunochemical techniques, such as the immunoblotting analysis with specific antibodies, may be especially valuable for further studies in this area.

ACKNOWLEDGMENT

We thank Mr. Joseph DeCicco for expert assistance. This work is supported in part by grants AG03853 from NIH, and #6-406 from March of Dimes Birth Defects Foundation; by grants from Muscular bistrophy Association, the Will Rogers Institute, and the Winifred Masterson Burke Relief Foundation.

REFERENCES

1. HoGEBOOM, G. H., and ScHNEIDER, W. C. 1953. Intracellular distribution of enzymes. XI. Glutamic dehydrogenase. J. Biol. Chern. 204:233-238. 2. DI PRisco, G., and GAROFANO, F. 1975. Crystallization and partial characterization of glutamate dehydrogenase from ox nuclei. Biochemistry 14:4673-4679. 3. DI 11usco, G., and CASOLA, L. 1975. Detection of structural difference between nuclear and mitochondrial glutamate dehydrogenases by the use of immunoabsorbents. Bio­ chemistry 14:4679- 4683. 4. KATo, G., and LoWRY, 0. H. 1973. Distribution of enzymes between nucleus and cy­ toplasm of single nerve cell bodies. J. Bioi. Chern. 248:2044-2048. 5. LEONG, S. F., and CLARK, J. B. 1984. Regional development of glutamate dehydrogenase in the rat brain. J. Neurochem. 43:106- 111. 6. FILLA, A., DE MICHELE, G., BRESCIA MoRRA, V., PALMA, V. , and CAMPANELLA, G. 1984. Clinical and biochemical studies on Friedreich's ataxia. ltal. J. Neurol. Sci. 2(Suppl. 4):47-54. SUBCELLULAR LOCALIZATION OF GDH 743

7. BERL, S., and CLARKE, D. D. 1984. The metabolic compartmentation concept. Pages 205-218, in HERTZ, L., KvAMME, E., McGEER, E. G., and ScHousBoE, A. (eds.), Glu­ tamine, glutamate and GABA in the central nervous system, Alan R. Liss, New York. 8. NICKLAS, W. J., NuNEZ, R., BERL, S., and DuvoiSIN, R. 1979. Neuronal-glial contri­ butions to transmitter amino acid metabolism: Studies with kainic acid-induced lesions of rat striatum. J. Neurochem. 33:839-844. 9. DENNIS, S.C., LAI, J. C. K., and CLARK, J. B. 1977. Comparative studies of glutamate metabolism in synaptic and nonsynaptic rat brain mitochondria. Biochem. J. 164:727- 736. 10. LAI, J. C. K., SHEU, K.-F. R., CLARKE, D. D., KIM, Y. T., and BLASS, J. P. 1985. Glutamate dehydrogenase in brain mitochondria and nuclei. Trans. Am. Soc. Neuro­ chem. 16:136. 11. LAI, J. C. K., SHEU, K.-F. R., CLARKE, D. D., KIM, Y. T., and BLASS, J.P. 1985. Is glutamate dehydrogenase a mitochondrial marker in brain? J. Neurochem. 44(Suppl.):S158. 12. LAI, J. C. K., and CLARK, J. B. 1979. Preparation of synaptic and nonsynaptic mito­ chondria from mammalian brain. Methods Enzymol. 55(Part F) 51-60. 13. LAI, J. C. K., SHEu, K.-F. R., and CARLSON JR., K. C. 1985. Differences in some of the metabolic properties of mitochondria isolated from cerebral cortex and olfactory bulb of the rat. Brain Res. 343:52-59. 14. LAI, J. C. K., and CLARK, J. B. 1976. Preparation and properties of mitochondria derived from synaptosomes. Biochem. J. 154:423-432. 15. SHEU, K.-F. R., LAI, J. C. K., and BLASS, J.P. 1984. Properties and regional distribution of pyruvate dehydrogenase kinase in rat brain. J. Neurochem. 42:230-236. 16. WEBER, K., and OsBORN, M. 1975. Protein and sodium dodecyl sulfate: Molecular weight determination on polyacrylamide gels and related procedures. Pages 179-223, in NEu­ RATH, H., and HILL, R. L. (eds.), The proteins, 3rd edition, Academic Press, New York. 17. SHEU, K.-F. R., and KIM, Y. T. 1984. Studies on the bovine brain pyruvate dehydro­ genase complex using the antibodies against kidney enzyme complex. J. Neurochem. 43:1352-1358. 18. SHEu, K.-F. R., LAI, J. C. K., KIM, Y. T., DoRANTE, G., and BAGG, J. 1985. Immu­ nochemical characterization of pyruvate dehydrogenase complex in rat brain. J. Neu­ rochem. 44:593-599. 19. LoWRY, 0. H., RosEBROUGH, N. J., FARR, A. L., and RANDALL, R. J. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chern. 193:265-275. 20. L0VTRUP-REIN, H., and McEWEN, B. S. 1966. Isolation and fractionation of rat brain nuclei. J. Cell Bioi. 30:405-415. 21. THOMPSON, R. J. 1973. Studies on RNA synthesis in two populations of nuclei in the mammalian cerebral cortex. J. Neurochem. 21:19-41. 22. AusTOKER, J., Cox, D., and MATHIAS, A. P. 1972. Fractionation of nuclei from brain by zonal centrifugation and a study of the ribonucleic acid polymerase activity in the various classes of nuclei. Biochem. J. 129:1139-1155. 23. CREE, P. Y., DAHL, J. L., and FAHIEN, L.A. 1979. The purification and properties of rat brain glutamate dehydrogenase. J. Neurochem. 33:53-60. 24. KING, K. S., and FRIEDEN, C. 1970. The purification and physical properties of glutamate dehydrogenase from rat liver. J. Biol. Chern. 245:4391-4396. 25. HERZFELD, A., FEDERMAN, M., and GREENGARD, 0. 1973. Subcellular morphometric and biochemical analysis of developing rat hepatocytes. J. Cell Biol. 57:475-483. 26. FRANKE, W. W., DEUMLING, B., ERMEN, B., JARASH, E. D., and KLEINIG, H. 1970. Nuclear membranes from mammalian liver. I. Isolation procedure and general char­ acterization. J. Cell Bioi. 46:379-395. 744 LAI ET AL.

27. DIPRisco, G., and STRECKER, H. J. 1970. Glutamate dehydrogenase of nuclear and extra- nuclear compartments of Chang's liver cells. Eur. J. Biochem. 12:483-489. • 28. PLAITAKIS, A., NICKLAS, W. J., and DESNICK, R. J. 1980. Glutamate dehydrogenase deficiency in three patients with spinocerebellar syndrome. Ann. Neurol. 7:297-303. 29. DuvoiSIN, R. C., CHOKROVERTY, S., LEPORE, F., and NICKLAS, W. 1983. Glutamate dehydrogenase deficiency in patients with olivopontocerebellar atrophy. Neurology (Cleveland) 33:1322-1326. 30. PLAITAKIS, A., BERL, S., and YAHR, M.D. 1984. Neurological disorders associated with deficiency of glutamate dehydrogenase. Ann. Neurol. 15:144-153. 31. SHEU, K.-F. R., CEDARBAUM, J., HARDING, B. J., CLARKE, D. D., and BLAss, J. P. 1985. Measurements of pyruvate and glutamate dehydrogenases as well as fumarase in platelet lysates from patients with Parkinson's and related diseases. VIII. Intl. Sym. Parkinson's Disease (Abstract): p38.